Materials Today: Proceedings xxx (xxxx) xxx

Contents lists available at ScienceDirect

Materials Today: Proceedings

journal homepage: www.elsevier.com/locate/matpr

A novel approach of various QAM with roll off factor variation using
raised cosine filter and SRRC filter for analysis of BER and SNR
Preesat Biswas a, Shanti Rathore a, M.R. Khan b
a
Department of ECE, Dr. C.V. Raman University, Bilaspur, CG, India
b
Department of ET&T Engineering, Government Engineering College, Jagdalpur, CSVTU, CG, India

a r t i c l e i n f o a b s t r a c t

Article history: In QAM is form of modulation which is generally used for modulating data signals into a carrier utilized
Received 9 November 2020 for digital communications. QAM, once used for digital transmission for wireless communications exhi-
Accepted 17 November 2020 bition is in a position to hold higher information rates concerning time than normal amplitude-
Available online xxxx
modulated schemes and phase-modulated schemes. This paper suggested A Novel Approach of Various
QAM with RoF Variation Raised Cosine Filter and Square root raised Cosine Filter for analysis of BER
Keywords: and SNR. The time delay between the raised cosine filter and Square root raised cosine filter. In addition,
SRRC
a comparative analysis of the proposed model with state-of-art is reported in terms of Bit Error Rate (BER)
RCF
RC
with a bit rate.
QAM Ó 2020 Elsevier Ltd. All rights reserved.
ISI Selection and peer-review under responsibility of the scientific committee of the Emerging Trends in
AWGN Materials Science, Technology and Engineering.
SNR
BER
RoF

1. Introduction 2. Qam scheme

In modern communication system, it is to necessary advance In the signals of modulation it preferred to many destination
technology used that should scope up with the development of point and how much far to communicate it. According this manner,
people and modern societies. Communication is a process which the carrier signal with modulation signal. Here in QAM having with
is source to destination communicate with the help of ‘‘transmit- two carrier signals are generated in-phase (‘‘I”) and quadrature
ting” and ‘‘receiving” of messages from that point. It should be (‘‘Q”) are get generated with a phase shift of 90o in between them.
transmitter signal as well as decode the received signal exact mes- Normally these are the baseband area. The output of two signal
sage those transmitted to transmitter [1]. In QAM with different which is first sum and then required to forward to needed RF sig-
roll-off factor (RoF) in RCF and SRRC. The effect of multipath prop- nal. This frequency conversion to requires amplification and
agation has become the major research field and it should be the demodulation is it as same reverse process manner in QAM. Here
necessary to pre-plan for of any system for designing multipath the QAM output is feed to the discriminator and divide into two
communication channels. In transmission of information, it gets half, each half is applied to a mixer.
in QAM (Quadrature amplitude modulation) reassemble of phase A local oscillator signal working as a mixer, one in phase and
and frequency. We ensure that reduced the band width till it other with 90o phase shift in a local carrier. The result of mixer
should be set to low value, which is pass through (See Fig. 1). has two output called ‘‘I” and ‘‘O”. Whereas ‘‘O” get generated
AWGN channel. In received signal which resolve phase and fre- 90o phase-shifted carrier is applied that time ‘I’ output get to
quency, demodulated, decoded signal, check the comparator, Error mixer. Output are 8 in ‘‘I” values and 8 in ‘‘Q” values, 256-QAM
count, and calculate BER(Bit error rate) with SNR(signal to noise with expectation of highly proficient and productiveness.
ratio).

https://doi.org/10.1016/j.matpr.2020.11.568
2214-7853/Ó 2020 Elsevier Ltd. All rights reserved.
Selection and peer-review under responsibility of the scientific committee of the Emerging Trends in Materials Science, Technology and Engineering.

Please cite this article as: P. Biswas, S. Rathore and M.R. Khan, A novel approach of various QAM with roll off factor variation using raised cosine filter and
SRRC filter for analysis of BER and SNR, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.11.568
P. Biswas, S. Rathore and M.R. Khan Materials Today: Proceedings xxx (xxxx) xxx

Fig. 1. Block diagram.

3. Raised cosine filter W  W0

a¼ ð6Þ
W0
In transmission systems, the bits are normally transmitted in The RoF indicates estimation of power use in RC emits within
the form of individual pulses. In order to regulate the power emis- the bandwidth W0. Therefore 0  a  1. Identify the sampling
sion within the ban-limit that is required. The signal has a major time, when we implement in RC pulses employed together with
lobe and minor lobe. Appearing minor lab, it tendency to to appear an equalizer at the receiver.
in the adjacent spectral band called as ISI. It to be nedder filters
system in the transmission due to avoid ISI. The best purpose hRC ðf Þ ¼ hTx ðf Þ:hc ðf Þ:hRx ðf Þ:hE ðf Þ ð7Þ
reduce the high attenuation is stop band filter. Alternatively, in
Where the transmission filter is hTx (f), the channel frequency
time domain pulse shaping is needed and RCF is the best choice
response is hc (f), receiver filter is hRx (f), the equalizer is hE (f). To
for this regard. The RRC is given by (5), and (6).
design the transmitter and receiver filter are
8
>
< h 1 i forjf j < 2W 0  W
jf jþW2W 0
hRC ðf Þ ¼ hTX ðf Þ:hRC ðf Þ ð8Þ
hRC ðf Þ ¼ cos p4
2
for2W 0  W < jf j < W ð1Þ
>
:
WW 0
and
0 forjf j > W
1
Where W-W0 is surplus bandwidth and from this signal Raised hE ðf Þ ¼ ð9Þ
hc ðf Þ
Cosine falls above bandwidth W0.So ‘‘Nyquist” pulse filter are
pulse that consequence no ISI at the sampling time. The Nyquist
pulse-shaping measure for zero ISI is satisfied. The following 3.1. Square root raised cosine filter
relation
xðkT c Þ ¼ dðkÞ ð2Þ The advanced approach of RCF i.e SRRC [9] filter is adopted here.
The SRRC filter, it generates a frequency reply with unity gain at
Where the low frequencies and complete at the higher frequencies. It is
It was simplified as generally used in communication systems. Whereas the transmit-
 ter to begin with applying the SRRC channel, and then the receiver
1 k¼0
dðkÞ ¼ ð3Þ section we have applies a matched filter.
0 k–0 The SRRC is given by (7) and (8)
  qffiffiffiffiffiffiffiffiffiffiffiffiffi
X
1
m
X fþ ¼ Tc ð4Þ hTX ðf Þ ¼ hRX ðf Þ ¼ hRC ðf Þ ¼ hSRRC ðf Þ ð10Þ
m¼1
Tc

Where X(f) is representing the Fourier transform signal x(t) and Tc 3.2. Additive White Gaussian Noise (AWGN)
the time period of the pulse.
cos½2pðW  W 0 Þt AWGN is caused due to electrons of thermal energy altogether
hRC ðt Þ ¼ 2W 0 sincð2W 0 tÞ ¼ 2
ð5Þ dissipative electrical components. It can be modeled as a a process
1  ½4ðW  W0Þt
which is Gaussian and consisting zero-mean. And random signal is
Another a is the RoF, it is determines the sharpness of the fre- that the collection of the random noise variable and a right away
quency response. from current signal as shown in (9) [3,4,10].
2
P. Biswas, S. Rathore and M.R. Khan Materials Today: Proceedings xxx (xxxx) xxx

Z ¼aþv ð11Þ
The Gaussian noise has probability distribution function repre-
sented as;

1 1 z  a2
pðzÞ ¼ pffiffiffiffiffiffiffi exp  ð12Þ
r 2p 2 r
The model of this noise assumes a PSD Gn(f) which is flat for all
the frequencies denoted as;
N0
Gn ðf Þ ¼ ð13Þ
2
According to [6] SRC pulses are very similar to the Nyquist
pulses which consists of finite bandwidth with PSD is

8
> T k f k 6 1 a
< n h c io 2T c

GSRRC ða; f Þ ¼ Tc
1 þ cos paT c k f k  1 a 1a
6 k f k 6 1þ a
>
:
2 2T c Tc 2T c

0 k f k > 1þ a
2T c

ð14Þ
Fig. 2. Time domain wave from of RC pulse shaping filter figure. Z 1
GSRRC ða; f Þdf ¼ 1 ð15Þ
1

The issue to indicate that the PSD could be a bilateral spectrum.

Such version noise is ubiquity to all digital communication systems
and is that the major noise source for many systems with physical
appearance of AWGN.
in almost every places AWGN in communication systems is
used to make a noise model which we have simulated here to
get appropriate estimation their performance.
h i h i
ð1þaÞ ð1þaÞ
4a cos Tc
p t þ Tc
4at
sin Tc
p t
sðtÞ ¼ pffiffiffiffiffi
2 ð16Þ
p Tc 1  4Tat
In time representation of SRRC pulses, the bandwidth is finite
and the value of the term ð1 þ Þ=Tc is the remaining same.

4. Result and discussion

4.1. Error rate calculation

This block helps to computing the error rate by making compar-

Fig. 3. Frequency domain wave from of RC pulse shaping filter. ison between received data and the delayed form of transmitted
data. The output of block is a three section or the components
which consists of firstly error rate secondly number of errors which

Fig. 4. Impulse response.

3
P. Biswas, S. Rathore and M.R. Khan Materials Today: Proceedings xxx (xxxx) xxx

Fig. 5. h1 RoF 0.35.

Fig. 6. h2 ROF = 0.45.

Fig. 8. Transmitted data and upsampled data.

sinðpt=TÞ
g ðt Þ ¼ t ¼ 1 to þ 1 ð17Þ
Fig. 7. Sample of h1 & h1xh2.
pt=T
We get result in ideal band limited to frequencies from  2T 1
Hz
to þ 2T
1
Hz with the help of raised cosine filtering for pulse With
has been detected and lastly the total number of symbols com-
the help of Matlab 2019a, plotted the time domain and frequency
pared which are compared [2].
domain illustration of the RCF for different values of a,i.e.a = 0.27,
a=0.35, a=0.45, a=1.
4.2. Transmit pulse shaping in Raised cosine filter We observed in time domain samples, that filter tail of the RCFr
with a greater than 0 dies down faster. As can be seen from the fre-
In transmit filtering using Nyquist pulse and formation symbol quency response, with a greater than 0, the filter response is ban-
am with a symbol period T can be transmitted without inter symbol dlimited only till  1þ
2T
a. We need a wider bandwidth to transmit the

interference (ISI) by using Nyquist pulse. wave from when compared to classical Nyquist bandwidth.
4
P. Biswas, S. Rathore and M.R. Khan Materials Today: Proceedings xxx (xxxx) xxx

Fig. 9. Transmitted data with alpha 0.35 & alpha 0.45.

Fig. 13. Impulse response in 256 QAM with a = 0.35 in 20 dB.

The combination of two signal does not coincide with the ordi-
nary channel since of it has a length. Increment span is equal to
constant, to get appropriate understanding between the reactions
and way better compliance with as shown in Fig. 7. This is ISI rejec-
tion capability of the RCF, and which is how to split the RCF filter-
ing between transmitter section and thath of the receiver section,
using raised cosine transmit and receive filter System.

Fig. 10. Transmitted data and square raised cosine. 4.3. Specifications of raised cosine filter

The most constraint of a RCF is its RoF in Fig. 8, which in a

roundabout way indicates the transmission capacity of the filter.
Perfect RC channels have an interminable number of taps. Subse-
quently, viable RC channels are windowed. The window length is
being controlled by means of utilizing the Channel Span in Symbols
property. We indicate the window length as six symbol lengths,
i.e., the channel ranges six image terms. This type of a channel
too encompasses [5] a bunch delay of three symbol terms. RCF
are used of pulse shaping, where the signal is upsampled. Hence,
we moreover ought to indicate the upsampling factor. We utilize
cosine transmit filter Framework object and set its properties to
get the required channel characteristics [5]. ‘‘This object plans a
direct- form of a typical polyphase FIR type filter which is consist-
Fig. 11. Transmitted data and square raised cosine.
ing energy as unity. The channel has an arrange of Nyquest sys-
4.2.1. Impulse responses of RRC and SRRC filters tem  samples PerSymbol, or Nyquest
Here we comparing with a typical RCF and a SRRCF. An ideal system  samplesPerSymbol + 1 taps”. We utilized the Gain prop-
normal RC pulse-shaping filter is equivalent to that of twice in erty to normalize the filter coefficients so that the sifted and unfil-
number of the ideal SRRCF in cascade configuration. Thus, the tered information matches when overlayed.
impulse response of an FIR filter Fig. 4 should be similar to that
of a RCF is being convolved with itself. We have Created a normal 4.3.1. Pulse shaping with RCF (raised cosine filter)
RCF with RoF 0.35 and 0.45, what is signifies that this filter has We [7] ‘‘create a bipolar information arrangement. Conse-
span 4 symbols with 3 samples per symbol. It was converted the quently, we utilize the RCF to profile the waveform without pre-
square-root filter with itself, then cut short the impulse response senting ISI” [5,7]. ‘‘The plot compares the digital data and the
outward from the greatest. So it has the equal length as h1 in inserted signal. It is troublesome to comparison of the two signals
Fig. 5 and h2 in Fig. 6. Normalize the response using the maximum. since the top reaction of the channel is deferred by the bunch delay
Then, we have compared the convolved square-root filter to the of the channel (Nyquest symbol/(2R)). We add (Nyquest symbol/2)
normal filter. zeroes after input  to level all the valuable tests out of the chan-

Fig. 12. Impulse response in 16 QAM, a = 0.35, 10 dB.

5
P. Biswas, S. Rathore and M.R. Khan Materials Today: Proceedings xxx (xxxx) xxx

Fig. 14. Frequency response.

Fig. 15. Eye diagram for in-phase signal and 16-QAM.

Fig. 16. Eye diagram for signal and 256-QAM. Fig. 17. Eye diagram for in-phase signal and 16-QAM.

6
P. Biswas, S. Rathore and M.R. Khan Materials Today: Proceedings xxx (xxxx) xxx

Fig. 20. Received signal, before and after in 256-QAM.

4.3.2. Roll-off factor(RoF)/Ratio

[7] ‘‘The step appears the impact that altering the RoF from on
and after 0.45 (blue bend) to 0.35 (ruddy remedy) has on the com-
ing about sifted output Fig. 11. The lower esteem for RoF results in
the filter to have a smaller move band resulting the sifted flag over-
shoot to be more prominent for red and blue bend”.

4.4. Square-root cosine filters (SRRCF)

We have done [8] ‘‘utilization of RCF is to part the sifting

between transmitter section and that of the receiver. In Both the
sections we utilized SRRC channels. The combination of transmit-
ter and receiver channel may be a RCF, which comes about in least
ISI”. We indicate SRRC filter by setting the shape as ‘Square root’’in
Fig. 18. Eye diagram for in-phase signal and 256-QAM. Fig. 12. [4] ‘‘The transmitted signal (maroon bend) is at that point
sifted at the receiver. With the help of usual unit energy normaliza-
tion guarantees that the pick up of the arrangement of the transmit
and receive channels is the same as the pick up of a normalized
RCF”. The sifted gotten flag, which is for all intents and purposes
indistinguishable to the signal sifted employing a single RCF, is
portrayed by the blue bend at the receiver-section shown in
Fig. 13.

4.5. Use pulse shaping on various QAM signal

Using different types of QAM into performance of pulse shaping

with the help of the matched filtering technique and by using a
combination of the SRRC filters [8,10]. ‘‘The raise cosine function
forms the filers and its results the enhancement of BER perfor-
mance can be evaluated in addition with forward error correction
(FEC) technique to communication line”. Adding FEC to the com-
munication connection to the pulse-shape filtering in different
types of QAM. i.e 16-QAM, 64-QAM, and 256-QAM. In communica-
tion link using, the binary data stream that comprises of a base-
band modulator channel, demodulator, pulse shaping and match
filters” [8]. ‘‘This exhibition a porting of the random data in a par-
ticular stem plot, presents the transmitted. Incoming signals in
received side constellation diagrams are created with a bit error
rate (BER)”.
Using the modulation technique in QAM with cosine filtering
and an AWGN channel, create RRC Filter parameters. Use the
‘Filter-Visualization-Toll’ in order to display the RRC filter
‘impulse-response’. Apply 256 QAM in zero dB power uses in
upsampling of the signal by the oversampling factor and applying
the RRC filter [11]. The ‘upfir’ function cushions the upsampled sig-
Fig. 19. Received signal, before and after in 16-QAM. nal with zeros at the conclusion to flush the filter”. At that point,
the function applies the channel [4]. By means of the number of
nel”. Comparing for the RCF group delay by deferring the input sig- k (bits-per-symbol) and the sps (number-of-samples-per – sym-
nal. Presently it is simple to identify how the ‘raised-cosine-filter’ bol), we have converted the ratio of energy per bit to noise ”EbNo”
up-samples and channels the signal at the input test times. This (power spectral density) to an SNR value use in awgn function.
happens the RCF capability to band constrain the signal whereas Then pass through the filtered signal through an AWGN channel”.
dodging ISI. Shape:- Normal with [8] RoF 0.45, Filter Span in Sym- Using [9] the received signal to downsample and filter the sig-
bols 6, Output Samples Per Symbol 8, and Gain is 1. nal, these downsample applying the same oversampling factor
7
P. Biswas, S. Rathore and M.R. Khan Materials Today: Proceedings xxx (xxxx) xxx

Table 1
.

S. No. No. of bits Roll-off Factor (RoF) 16-QAM Error 64-QAM Error 256-QAM Error
1 3*10^5 a = 0.35 0.00179 536 0.02640 7929 0.07870 23,599
2 9*10^5 a = 0.35 0.00810 542 0.02640 7916 0.07860 23,593
3 3*10^5 a = 0.45 0.00171 1535 0.02630 23,650 0.07880 70,932
4 9*10^5 a = 0.45 0.00170 1529 0.02640 23,732 0.07860 70,764
5 3*10^5 a = 0.35 0 0 0.000760 228 0.02020 6149
6 9*10^5 a = 0.35 0 0 0.000793 238 0.02000 5987
7 3*10^5 a = 0.45 0 0 0.000772 695 0.02030 18,291
8 9*10^5 a = 0.45 0 0 0.000760 684 0.02020 18,145
9 3*10^5 a = 0.35 0 0 0 0 0.000617 185
10 9*10^5 a = 0.35 0 0 0 0 0.000560 168
11 3*10^5 a = 0.45 0 0 0 0 0.000612 551
12 9*10^5 a = 0.45 0 0 0 0 0.000587 528

Fig. 21. 16-QAM before after convergence.

Fig. 23. 256-QAM before after convergence.

4.6. Eye pattern the filter effects

In an eye pattern Figs. 15 and 16 the filter effects of 16-QAM,

256-QAM at first we reduced the Eb/N0 s value editing and regen-
erated the received data in. At cascade a RCF as a matched pair of
filters 256-QAM, 1024-QAM as shown in Figs. 17 and 18. In this
system due to a high value of SNR of the signal without having
any type of multipath effects, ISI reduction output data in matched
pair of pulse shaping RRC filter and it is not consisting zero–ISI
until the second RRC filter are connected. The noise eye pattern sig-
Fig. 22. 64-QAM before after convergence.
nal after the matched receive filtering the pulse shaping and show
ISI as a narrowing.
applied for upsampling the transmitted signal. Filtering by utilizing The constellation diagram of that of the received signal with
the same SRRC filter functional to the transmitted signal. Applying applying filtering and without applying filtering is shown Figs. 19
SRRC filter, we know that in every filtering operation comprises of and 20. Scaling factor of the received signal by the square root of
delay and that delays in addition to the signal by half of the filter the number of samples per symbol to normalize the transmit and
length in symbol. Total delay from transmitted signal which is receive power levels.
come from received filtering equal the filter length we tried to kept
the same size as far as possioblend and after that we interpret for
the delay between the transmitted and received signal. After 4.7. After SRRC filter using coars and fine synchronization correct
removing we have removed the first length of filter symbols into phase and frequency offset in 16-QAM, 64-QAM and 256 QAM
the decimated signal to account for the collective delay of the
transmit and receive filtering actions. We Eliminate the last filter 4 Reassemble the Coars signal of significant phase and fre-
symbols in decimated signal to certify the number of samples in quency offsets in 16-QAM, 64-QAM and 256-QAM signal. First, cor-
the demodulator output contests of the number of samples in rect the coars frequency offset using the estimate provided by the
the modulator input [4]. By this QAM demodulation function we coars frequency compensator. Correction and tuned properly help
evaluated the demodulation process for the received filtered signal of carrier synchronizer, because it was set in low value within
demodulate to bit function to convert the recovered integer sym- the normalized bandwidth. The received signal applying phase
bols into binary data and perform calculation for the number of and frequency offset corrections and resolve phase ambiguity
errors and the associated BER. using the preambles.
8
P. Biswas, S. Rathore and M.R. Khan Materials Today: Proceedings xxx (xxxx) xxx

Table 2

S. No. QAM Sync Data TtlErr Sync Data BER Phase offset in degrees resPhz TtlErr resPhz BER
1 16-QAM 16,116 0.5042 90.1401 5 0.0001564
2 64-QAM 23,874 0.4979 90.3393 5 0.0001043
3 256-QAM 31,674 0.4955 90.4005 560 0.0088

analysis and emphasized that ISI degrades the overall performance

of the system if ISI is not eradicated before transmission of digital
signal.

CRediT authorship contribution statement

Preesat Biswas: Conceptualization, Methodology, Software,

Data curation. Shanti Rathore: Visualization, Investigation, Writ-
ing - original draft. M.R. Khan: Supervision, Software, Validation,
Writing - review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing finan-

cial interests or personal relationships that could have appeared
Fig. 24. Phase direction in different QAM.
to influence the work reported in this paper.

References
5. Discussion of results
[1] C.H. Cheng, IEEE Signal Process Mag. 23 (1) (2006) 88–96.
In transmission side without filtering the Random bits and fre- [2] S. Sharifian, Masters of Sciences Dissertation, Chalmers University of
quency with RoF alpha = 0.35, alpha = 0.45 in power 10 dB in Figs. 2 Technology, 2010.
[3] C.K. Madsen, G. Lenz, IEEE Photonics Technol. Lett. 10 (1998) 994–996.
and 3. Using normal impulse response in symbol model using [4] J. Vojtech, M. Karasekt, J. Radill, ICTON 19 (2007) 71–75.
alpha = 0.35, alpha = 0.45 Figs. 4–6 and pass through square-root [5] Tulsi Pawan Fowdur, Louvi Doorganah, Performance of modified and low
filter get result shown symbol in Fig. 7. The unity gain energy of complexity pulse shaping filters for IEEE 802.11 OFDM transmission, J. Inform.
Telecommun. 3 (3) (2019) 361–380, https://doi.org/10.1080/
Nyquest with Symbol Per sample, which roll-off factor
24751839.2019.1586360.
alpha = 0.35, alpha = 0.45 pass through narrow transition in Figs. 8– [6] William H. Tranter, K. Sam Shanmugan, Theodore S. Rappaport, Kurt L. Kosbar,
11. Both the transmitter and receiver employ filter is a RRCF which Principles of Communication Systems Simulation with Wireless Applications,
is minimum ISI in Figs. 12 and 13. Pulse Shaping on QAM Upper Saddle River, NJ, Prentice Hall, 2004.
[7] F. Khair, H.P. Fakhriy, I.W. Mustika, B. Setiyanto, S.M. Idrus, Modeling and
(M = 16,256) with alpha = 0.35 power using 10 dB and 20 dB simulation of OFDM scheme for radio over fiber (RoF), in: 2015 2nd
shown in Figs. 12 and 13 and determine frequency response in International Conference on Information Technology, Computer, and
Fig. 14. Scatter Plot in phase with different type of QAM (M = 16, Electrical Engineering (ICITACEE), Semarang, 2015, pp. 376–381, doi:
10.1109/ICITACEE.2015.7437833.
& 256). In Cascade eye pattern of different type of QAM (M = 16, [8] N.S. Alagha, P. Kabal, Generalized raised-cosine filters, IEEE Trans. Commun. 47
& 256) shown in Figs. 15–18. Constellation in received signal (7) (July 1999) 989–997. doi: 10.1109/26.774849.
before and after filtering as shown in Figs. 19 and 20. In Table 1we [9] G. Braileanu, Digital filters with implicit interpolated output, IEEE Trans. Signal
Process. 45 (10) (Oct. 1997) 2551–2560, https://doi.org/10.1109/78.640718.
have shown the different RoF, alpha = 0.35 and alpha = 0.45 with [10] I.A. Shaheen, A. Zekry, Complexity analysis of FBMC/OQAM transceiver using
bit error rate (BER), and we have also compared the occasioning SRRC polyphase filter, in: 2019 International Conference on Promising
errors in the receive signal with adapting different noise levels, Electronic Technologies (ICPET), Gaza City, Palestine, 2019, pp. 67–72. doi:
10.1109/ICPET.2019.00020.
since SNR in 10 dB then increasing 15 dB, 20 dB this ratio causes [11] H. Sari, G. Karam, V. Paxal, K. Maalej, Trellis-coded constant-envelope
less bit errors and decreasing in this ratio causes higher bit errors modulations with linear receivers, IEEE Trans. Commun. 44 (10) (Oct. 1996)
rate. After filtering QAM Coars are tuned into Synchronization Cor- 1298–1307, https://doi.org/10.1109/26.539770.
rect Phase and Frequency Offset Shown in Figs. 21–23. Phase angle
are arranged in Table 2 shown different QAM with synchronous Further Reading
data iteration. Synchronous Data BER, Phase Offset, rephase itera-
[1] G. Keiser, Optical Fiber Communications, third ed., McGraw-Hill, 2000.
tion and rephase with BER and phase direction in different QAM [2] Y. Chaba, R.S. Kaler, OPTIK 121 (2010) 813–817.
shown in Fig. 24. [3] W.H. Loh, R.I. Laming, N. Robinson, A. Cavaciuti, F. Vaninetti, C.J. Anderson, M.
N. Zervas, M.J. Cole, IEEE Photonics Technol. Lett. 8 (1996) 944–946.
[4] M. Liu, B. Li, W. Tang, A new method for MQAM signals identification, in: 2010
6. Conclusion 6th International Conference on Wireless Communications Networking and
Mobile Computing (WiCOM), Chengdu, 2010, pp. 1–4. doi: 10.1109/
A Novel Approach of Various QAM with RoF Variation RCF and WICOM.2010.5600729.
[5] N.S. Alagha, P. Kabal, Generalized raised-cosine filters, IEEE Trans. Commun. 47
SRRC filer for analysis of BER and SNR have been simulated with (7) (July 1999) 989–997, https://doi.org/10.1109/26.774849.
and without using the filtering operation for baseband transmis- [6] V. Fung, T.S. Rappaport, Bit-error simulation of pi /4 DQPSK in flat and
sion and it is observed that the error rate depends on the modula- frequency-selective fading mobile radio channels with real time applications,
in: ICC 91 International Conference on Communications Conference Record,
tion scheme used, the ratio (Eb/No) and stipulations of the Denver, CO, USA, 1991, pp. 553–557, vol. 2. doi: 10.1109/ICC.1991.162425.
channels, ISI. This paper has analyzed the performance of every [7] N. Voudoukis, Performance analysis, characteristics, and simulation of digital
system for a number QAM over the AWGN. From our various oper- QAM, Eur. J. Electr. Eng. Comput. Sci. 1(1). DOI:https://doi.org/10.24018/ejece.
2017.1.1.3.
ation and methodology used we found that without filter is com-
patibly less efficient than that of model with filter, which